When one of the world’s largest pharmaceutical companies commits up to a billion dollars to deploy agentic AI across its entire business, it signals something has fundamentally shifted in how enterprises think about this technology.

Merck and Google Cloud announced at Cloud Next 2026 in Las Vegas on April 22 a multi-year partnership valued at up to $1 billion. The deal will see an agentic AI platform built on Google Cloud’s Gemini Enterprise deployed across Merck’s research and development, manufacturing, commercial, and corporate functions. Google Cloud engineers will work directly alongside Merck teams for what the companies described as a collaboration expected to span at least a decade.

This is not a software subscription or a pilot program. It is a structural commitment to rebuild how a major enterprise operates.

What Merck Is Actually Building

The scope here matters. Merck is not applying AI to one department or automating a single process. The partnership covers four distinct business domains:

R&D — the most complex and high-stakes environment in pharma, where data volumes are enormous and the cost of mistakes is measured in years and billions.

Manufacturing — where consistency, quality control, and throughput are non-negotiable, and where AI agents can monitor, flag, and adjust in ways humans cannot at scale.

Commercial — including how Merck goes to market, manages customer relationships, and drives revenue growth.

Corporate functions — the operational backbone: finance, HR, legal, procurement, and everything that keeps a 60,000-person organisation running.

Each domain presents different challenges, different data types, and different risk tolerances. Building a unified agentic platform that works across all of them is genuinely difficult work.

The Scale of the Signal

A $1 billion investment is significant even for a company of Merck’s size. But the more revealing detail is the expected decade-long timeline. That is not the language of experimentation. That is the language of infrastructure.

Merck is treating agentic AI the same way enterprises treated cloud migration a decade ago: not as a discrete project with a completion date, but as an ongoing transformation of how the company operates at a fundamental level.

The embedded model also stands out. Google Cloud engineers working inside Merck is similar to how large technology consulting firms traditionally partnered with enterprises on core systems implementations. It reflects a recognition that deploying AI agents at this scale requires sustained, hands-on expertise — not just access to a platform.

What This Means for Business

Most companies are not Merck. But the pattern here is instructive for any business weighing its own AI strategy.

First mover advantage in your industry is still real. Merck is building AI-powered workflows into its core operations now. Competitors who delay are not staying neutral. They are falling behind. The same dynamic applies at every scale.

Agentic AI requires a data foundation. Deploying agents across R&D, manufacturing, and commercial operations only works if the underlying data is clean, connected, and governed. Companies that have invested in data infrastructure are positioned to move fast. Companies that have not will hit that wall first.

Point solutions are a dead end. The Merck approach is explicitly cross-functional. Agents that cannot share context across departments create the same silos that plagued legacy software. The value compounds when the system is coherent.

Implementation requires real expertise. The embedded team model Merck and Google Cloud are using exists precisely because deploying AI across complex operations is not something you configure over a weekend. Business owners evaluating AI need partners who understand both the technology and how to apply it to specific operational contexts.

For businesses that are ready to move beyond experimenting and start deploying AI that actually changes how work gets done, the question is not whether this is the direction. Merck’s $1 billion bet makes that clear enough. The question is how to get there from where you are today.

Enterprise DNA put together a free field guide on exactly this: the full Claude ecosystem, Claude Code, and how to roll agents out without breaking things. Get the guide.